How the filtration flow rate affects the filter internals and what happens to the particles in the different scenarios?
Believe it or not, the flow rate, especially in the pressure leaf filter, is a very critical point. Flow rate control must be continuously improved along the rest of the filtration system.
The filter cake, the pump curve and the pressure differential express about the flowrate. Understanding what is going on inside the filter can give us a good idea of what may be causing a problem and how to solve it.
What is the flow rate?
Flowrate is a scale that means how much liquid is going through a certain area of mesh in certain time, its regularly measured by litres per hour per square meters (l/h*m²). This is a very important information while calculating the filter size at the design phase of the filtration system.
Is important to notice that this rate is not fixed, it variates with the liquid properties as viscosity, expected percentage of solids, nature of the solids and the addition of filter aid. Any of these variables can affect the average ideal flow rate, – I mention ideal flow rate, because at the designed phase we use an average media flow rate. This media comes from already existing systems known to work properly in the market -.
Nevertheless, flowrate would be adjusted along the first filtration cycles when a new system is started up. The reason is that the input in to the process will never be the same. (that’s why in new designs we recommend the implementation of flow driven pumps system, so the flow would be monitored and adjusted as is needed (as the differential pressure increases)).
So, now that you know this. We can only tell that at our design phase we use an average flow rate from the collection of our experience in the market. For example, let use liquid sulphur as example, with an average 0.3% solids and including the lime dosing for acid neutralization, the ideal flow rate would be 680 l/h*m². This ideal flow rate would allow a good cake formation that would have low sulphur content and easier cake release.
How does internal flow work? : Internal flow and cake formation
Let’s take as example: Pressure leaf filter with a horizontal vessel and vertical leafs. Let assume a filtration process of liquid sulphur (with the previous specifications mentioned before) and cellulose pre-coat.
Starting by the point of filling the filter: the filter will be empty, it must be filled with clean liquid sulphur that will come from clean storage, pre-coat tank or from the running filter if there are 2 filters in parallel. For this, we open the vent while keeping the outlet closed. Once the filter is full of sulphur, we can close the vent, start the pumping from the pre-coat mixture to build a bit of positive pressure and, finally, open the outlet in recirculation. Is important to mention that the flow rate of the cellulose precoat layer building should be higher than the main filtration flow rate, this is because we want to create an even and neat pre-coat layer. Cellulose material is open by its nature, therefore higher flow will make a strong layer without making it too tight.
Whether the flow rate in the pre-coat process is higher the principle is the same as in the main filtration. At this point we have liquid sulphur with a 5% average concentration of cellulose coming in to the filter, these particles should spread in the filter and distribute all around the free space, at the same time we have a good amount of cellulose particles getting caught by the mesh covering the filter leafs as the liquid passes though.
At the beginning, the bigger particles will get retain first. Over the time, creating another layer with a lower porosity on top of the already existing filter mesh will make that even smaller particles will be caught by this new layer creating a new one with even lower porosity. Don’t imagine this a layer of paper pilling one in top of other, is more like a continuous random spread of particles all around. Most of the time, particles will go through all the formed layers, recirculating until they can go through and are trapped. That’s why we need certain time of recirculation until these have all found a place to settle.
When the flowrate is ideal, the cake should be an even layer of cellulose material all covering the mesh and this same principle applies to the main filtration.
When you have a proper pre-coat layer and you start the main filtration with an adequate flow rate, the particles will behave in the same way, except from the fact that there won’t be any particles passing thought from the beginning, the cellulose pre-coat layer works as a barrier. Every particle of the main filtration will be retained at some level at the precoat layer.
But by the contrary there are other different scenarios where the pre-coat is not well formed. These could be visible just after the precoat layer cake shape or by the final filtration cake shape.
Let’s take as first example a too low flowrate. With a low flowrate the problem starts as soon as we start filling the tank with the cellulose concentration, the slow flow won’t carry the particles evenly in the filter, giving the particles the chance to settle. This does not mean that they will be all in the bottom of the vessel, but particles will not be distributed evenly over the filter leaf.
What happens is that the particles start going through the mesh from the bottom up. So, instead of forming a layer evenly all around the mesh, the layer will start to grow from the bottom. The particles won’t reach a higher level on the mesh before they cannot get caught easily in a lower level (this is not a rule). Therefore, there is cake being build all around but not at the same rate. The result of this is a precoat layer that resembles a drop. A big concentration of precoat at the bottom and a thin one and sometimes is not present at the top of the leaf
As you can imagine this means a lot of problems, even If you have a good flow rate in the main filtration, this bad cake formation of the precoat will give you problems. (lets call this BS-P-D for “Bad scenario – Pre-Drop”)
Our second example: let’s take a pre-coat cake that is being built with a too high flowrate, in this scenario the problem does not start right away. When the cellulose mix come in to the filter vessel, particles will spread well all over the available area in the filter, the problem starts when the layers formation begins.
The particles that are being caught (retained) , create the first layer, only the big particles will remain in the course filter mesh, remember that these will create a new layer of a different but lower porosity, consider that the new layer is created by uneven length and thickness of fibres, (close attention from here) therefore every time a fibre gets retained, it decreases the porosity of it surrounding area, this happens in a different grade that other fibre will do if it gets caught in a similar distance away from the filtration mesh. (Please read it/see it again).
So, having this in mind, while in an ideal flow rate, a fibre would have certain freedom to flow around the cake trying to find an area to go through until it gets retained. In a too high flowrate, fibres will be pushed against any area with enough permeability to allow flow, considering this, is not hard to imagine that some areas in the pre-coat layer will have so many particles being pushed against each other that they will soon create a very low permeability in those areas, resulting in a partial blockage.
We will recognise this because the cake will be formed with an uneven spread layer that look like the sea surface, there are small sunken areas all over the precoat layer, along the waves, thinner area, have a lower permeability (if not completely blocked). Therefore, affecting the nominal filtration area that should have available with the precoat layer. (let’s call this BS-P-W for “Bad Scenario – Pre-waves) we will analyse further how this would affect the 3 scenarios of main filtration.
If you have any question or you would like to make your filtration system perform better, don’t hesitate to contact us at firstname.lastname@example.org or using the contact form in the website. We will be happy to help you.
Autor: Alfredo Rapetta
Co-Autor: Jan Hermans